Using the latest Docker image stable-5390-3 on an AWS c5.xlarge server.
Near the beginning of a meeting, I have a period of about one or two minutes in which the audio is extremely bad, like cut for two second, work for a fraction of second, cut again etc. During that time the video is fine. Once that period is over, the audio get back to normal and then stay normal for the rest of the meeting.
It’s not clear what trigger that period, but often it is when talking continuously for 10s/20s then stop completely. The awful audio period start just after that.
I tried to redirect the rtmp stream to youtube, an nginx/rtmp server, or even just a ffmpeg acting as a server and get the same issue.
Nothing is standing out to me in the log log.0.txt or ffmpeg.0.txt during that period.
Does anybody had a similar issue ? Any advice how to debug it more thoroughly ?